Comparison of initial HRCT features of COVID-19 pneumonia and other viral pneumonias

Background: Multicenter retrospective comparison of the rst high-resolution computed tomography (HRCT) ndings of coronavirus disease 2019 (COVID-19) and other viral pneumonias. Methods: We retrospectively collected clinical and imaging data from 254 cases of conrmed viral pneumonia in 20 hospitals in Yunnan Province, China, from March 1, 2015, to March 15, 2020. According to the virus responsible for the pneumonia, the pneumonias were divided into non-COVID-19 (133 cases) and COVID-19 (121 cases). The non-COVID-19 pneumonias included 3 types: cytomegalovirus (CMV) (31 cases), inuenza A virus (82 cases), and inuenza B virus (20 cases). The differences in the basic clinical characteristics, lesion distribution, location and imaging signs among the four viral pneumonias were analyzed and compared. Results: Fever and cough were the most common clinical symptoms of the four viral pneumonias. Compared with the COVID-19 patients, the non-COVID-19 patients had higher proportions of fatigue, sore throat, expectorant and chest tightness (all p<0.000). In addition, in the CMV pneumonia patients, the proportion of patients with combined acquired immunodeciency syndrome (AIDS) and leukopenia were high (all p<0.000). Comparisons of the imaging ndings of the four viral pneumonias showed that pulmonary lesions of COVID-19 were more likely to occur in the peripheral and lower lobes of both lungs, while those of CMV pneumonia were diffusely distributed. Compared with the non-COVID-19 pneumonias, COVID-19 pneumonia was more likely to present as ground-glass opacity (GGO), intralobular interstitial thickening, vascular thickening and halo sign (all p<0.05). In addition, in the early stage of COVID-19, extensive consolidation, brous stripes, subpleural lines, crazy-paving pattern, tree-in-bud, mediastinal lymphadenectasis, pleural thickening and pleural effusion were rare (all p<0.05). Conclusion: The HRCT ndings of COVID-19 pneumonia and other viral pneumonias overlapped signicantly, but many important differential imaging features could still be observed.


Introduction
Coronavirus disease 2019 (COVID-19) broke out in Wuhan, China, in December 2019; the disease is highly infectious, has a long incubation period and is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [1,2] . COVID-19 spreads at an alarming rate. As of April 20, COVID-19 has spread to 211 countries worldwide, with more than 2.3 million con rmed patients and 164,976 deaths. Currently, COVID-19 has become an inevitable global health crisis.
Early studies have shown that almost all COVID-19 patients have pneumonia [3,4] . However, at this time of year, pneumonias caused by other pathogens are also common [5][6][7] . Therefore, in this COVID-19 pandemic, the differential diagnosis of different viral pneumonias is di cult but is also important. Realtime reverse transcription-polymerase chain reaction (RT-PCR) is the gold standard for the diagnosis of viral pneumonia. However, recent reports have shown that the sensitivity of RT-PCR for the detection of COVID-19 is as low as 60-71% [8] , and the high false negative rate limits the rapid identi cation of viral pneumonia by RT-PCR.
Currently, computed tomography (CT) can play an important role in the diagnosis and treatment of viral pneumonia [9][10][11] . Studies have shown that the typical CT ndings of COVID-19 include ground-glass opacity (GGO) and partial consolidation in the peripheral areas of both lungs with a round shape and without cavities, pleural effusion and lymphadenopathy [12,13] . After treatment, chest CT can detect the dynamic changes in COVID-19 [14,15] . Other studies have shown that the imaging ndings of viral pneumonias and bacterial pneumonias are different [9,10] . However, little is known about the differences in imaging ndings between COVID-19 and other viral pneumonias.
Therefore, the purpose of this study was to clarify the basic clinical features and potential differences in high-resolution computed tomography (HRCT) ndings of COVID-19 and other viral pneumonias.

Patients
This retrospective study was approved by our institutional review board, and patient consent was waived. This multicenter study retrospectively analyzed patients who underwent chest CT for suspected viral pneumonia due to fever, fatigue or respiratory tract symptoms at 20 hospitals in Yunnan Province, China, from March 1, 2015, to March 15, 2020. RT-PCR assays were performed to identify in uenza A virus, in uenza B virus, respiratory syncytial virus, parain uenza virus, adenovirus, SARS coronavirus, SARS-CoV-2, Epstein-Barr virus, measles virus, and other viruses from nasopharyngeal swabs or bronchoalveolar lavage uid. The study only included pneumonia patients infected with one virus, and patients with multiple respiratory viruses or bacterial or fungal infections were excluded. A total of 351 viral pneumonia patients were diagnosed in this study. The further selection process for viral pneumonia patients is shown in Fig. 1. According to the clinical guidelines for COVID-19 [15] , among the 121 COVID-19 patients, 22 were mild, 76 were moderate, 20 were severe, and 3 were critical. In addition, the study excluded one case of Epstein-Barr virus, two cases of herpes virus, one case of measles virus, two cases of varicella virus and one case of mumps virus. All patients were admitted within 4-7 days after the onset of acute symptoms. The patient's age, sex, history of exposure, and clinical symptoms (fever, cough, fatigue, dyspnea, sore throat, runny nose, expectoration, headache, muscle aches, chest tightness, chest pain, nausea and vomiting, diarrhea and no obvious symptoms), underlying diseases (hypertension, diabetes, coronary heart disease, liver disease, tumor, acquired immunode ciency syndrome (AIDS) and leukopenia), hospital admission time and CT examination time were recorded. Patients completed the chest CT examination within 48 hours after admission. According to the virus found in the lungs, the patients were divided into four groups: cytomegalovirus (CMV), in uenza A virus, in uenza B virus and COVID-19. The number of cases from each hospital is shown in Table 1. All patients were from Yunnan Province, China.

Ct Protocol
HRCT examination: CT scanners with 16 or more detector rows (Siemens, Germany; Philips, the Netherlands; and GE, USA) were used. The patient was scanned in supine position while holding his or her breath after inhalation. The scanning range was from the thoracic inlet to the costophrenic angles. Scanning parameters: detector collimation width 64 × 0.6 mm or 128 × 0.6 mm, tube voltage 120 kV, adaptive tube current, high-resolution algorithm reconstruction, reconstruction layer thickness 1 or 1.5 mm and layer spacing 1.5 mm.

Chest Ct Analysis
Three Chinese radiologists were blinded to the RT-PCR results, all patient information, and type of viral pneumonia. First, two experienced radiologists (Yilong Huang and Yuanming Jiang) in the cardiothoracic group independently read the radiographs. When their opinions were inconsistent, they discussed them and reached a consensus, which was reviewed and con rmed by the third senior radiologist in the cardiothoracic group (Bo He). The morphological signs of the rst CT examination after admission were analyzed. The CT imaging evaluation [9,10,12,17] included lesion distribution (peripheral, central), location (left upper lobe, left lower lobe, right upper lobe, right middle lobe and right lower lobe) and signs (GGO [ground-glass opacities], partial consolidation, multifocal consolidation, focal consolidation, brous stripes, septal thickening, intralobular interstitial thickening, subpleural lines, crazy-paving pattern, tree-inbud, bronchial wall thickening, bronchiectasis, vascular thickening, air bronchogram, halo sign, mediastinal lymphadenectasis, pleural thickening and pleural effusion). The window width and level were set to 1600/-600 HU.
Statistical analysis SPSS 25.0 software was used for statistical analysis. Count data are expressed as frequency, and measurement data are expressed as` x ± s. One-way analysis of variance (ANOVA) was used for age, which had a normal distribution, and the least signi cant difference (LSD) method was used for pairwise comparison. The distribution, location and signs of pulmonary lesions in different viral pneumonias were compared using χ2 or Fisher's exact probability method. The Z-test (Bonferroni method) was used for pairwise comparisons. p < 0.05 was considered statistically signi cant. Five patients with in uenza A had a history of poultry contact. Among the four viral pneumonias, fever and cough were the most common clinical symptoms, and the highest proportion was found in the in uenza A patients (p < 0.000). Compared with the COVID-19 patients, the non-COVID-19 patients had higher proportions of fatigue, sore throat, expectoration and chest tightness, while the COVID-19 patients were more likely to be asymptomatic, and the differences were statistically signi cant (all p < 0.000). Compared with the other viral pneumonia patients, the CMV pneumonia patients had higher proportions of AIDS and leukopenia (all p < 0.000). Table 2 shows the clinical characteristics of the included patients.

Distribution and location of the pulmonary lesions of COVID-19 and other viral pneumonias
Thirty-three COVID-19 patients were negative on the rst CT examination. Figure 2

Chest CT ndings of viral pneumonia
The rst CT images of pulmonary lesions in CMV, in uenza A virus, in uenza B virus, Non-COVID-19 and COVID-19 were compared (  Continues data are expressed as mean ± SD, and categorical data are presented as n (%). *, compared with COVID-19, signi cance was considered when P < 0.05.

Discussion
COVID-19 is threatening human health and safety worldwide. Considering the similarity in outbreak time and clinical manifestations between COVID-19 and other viral pneumonias, this study systematically analyzed the differences in rst chest CT ndings between COVID-19 and other viral pneumonias. Our study found that although it is di cult to completely differentiate COVID-19 from other viral pneumonias, COVID-19 still has some unique CT features.
This study included 254 patients with con rmed viral pneumonia. The main clinical manifestations of all patients were fever and cough. However, the incidence of fever and cough in COVID-19 patients was low, which was consistent with the results of Zhao et al [18][19] . This may be related to the low virulence of SARS-CoV-2. In this study, 15 patients (12.40%) showed no obvious clinical symptoms, and 33 patients (27.27%) had negative CT ndings, which was consistent with the results of Guan et al. [20] . However, the patients' RT-PCT results were still positive for SARS-CoV-2, indicating that they were still infectious and should be isolated for observation and receive antiviral treatment. It is worth noting that patients with AIDS and leukopenia were more prone to developing CMV pneumonia.
The periphery of the lower lobes of both lungs was the most common area of lesions in COVID-19 and in uenza pneumonias, which is consistent with the results of previous studies [17] . However, we found that a high proportion of COVID-19 lesions occurred in the peripheral area, while CMV pneumonia was usually diffusely distributed in both lungs. This study systematically analyzed and compared the CT ndings of COVID-19 and other viral pneumonias. We found that GGO was more common in COVID-19 than in other viral pneumonias, and multifocal consolidation was more common in other viral pneumonias, a result that was consistent with previous studies [17,19] . To the best of our knowledge, studies comparing microscopic signs between COVID-19 and other viral pneumonias are still rare. Intralobular interstitial thickening, vascular thickening and halo signs were more likely to occur in COVID-19 than other viral pneumonias, and brous stripes, subpleural lines, crazy-paving pattern, tree-in-bud, pleural effusion and mediastinal lymphadenectasis occurred less frequently in COVID-19 than in other viral pneumonias. This may be related to the slower development of COVID-19 than other viral pneumonias [19] . It is worth noting that tree-in-bud was observed in 35.29% of in uenza pneumonia cases, and similar results were observed by Shiley et al. [21] , but almost no tree-in-bud was observed in COVID-19.
Compared with other viral pneumonias, CMV pneumonia often presents as characteristic bronchiectasis, and immunocompromise is an important risk factor. In previous studies, it has been called "AIDS associated bronchiectasis" [22] . With this information combined with the clinical data, radiologists can better identify the type of viral pneumonia [23] .
This study has some limitations. First, although we tried our best to collect the clinical and imaging data of patients with viral pneumonia in Yunnan Province, the number of con rmed cases of adenovirus, measles virus, herpes virus and other viruses was relatively small, but the differences in imaging signs among them are very interesting and will be studied in the future. Second, most of the included CMV pneumonia patients had AIDS, which is likely to cause a selection bias regarding other CMV pneumonia patients. Third, different populations, such as infants, children and elderly adults, may be susceptible to different viruses, and their signs of pulmonary lesions on imaging might be different. Because of the sample size, we did not perform a subgroup analysis for age. In future studies, more efforts should be made to determine the differences in the imaging characteristics of different populations.
In summary, the analysis and comparison of the chest CT ndings of COVID-19 and other viral pneumonias showed that the chest CT ndings partially overlapped, but many signi cant imaging features could still be observed, which is helpful for the early differential diagnosis of COVID- 19  Availability of data and materials: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Con icts of interest: The authors have no con icts of interest to declare in relation to this article.
Funding: No funding was provided for this study.